Cognoa’s AI platform for autism diagnosis gets first FDA stamp


Cognoa has gained regulatory recognition for its machine learning software as a class II diagnostic medical device for autism — meaning the digital health startup is now positioned to submit an application for full FDA clearance.

It’s a first but important regulatory step for a business that was founded back in 2014, and plays in a still nascent digital health space where untested ‘wellness’ apps are far more plentiful than medical technologies with robust data to prove out the efficacy of their interventions.

Discussions with the FDA started in early 2017, says Cognoa CEO Brent Vaughan, adding that it’s hoping to gain full FDA clearance this year.

He says the ultimate goal for the US startup is to become a standard part of domestic health insurance-covered medical provision — and for that FDA clearance is essential to opening the doors.

We first covered the Cognoa at launch in 2014 and the following year when it was still being careful to describe its technology as a screening rather than a diagnostic system.

It’s since gathered enough data to be confident in using the ‘D’ word — having run a pilot with 250,000 parents, offering free screening for their children so it could gather more data to refine its machine learning models.

“We were lucky that we had investors,” says Vaughan. “There’s not a huge business model in providing free screening services to kids, right, because we were certainly never going to sell ads. That wasn’t the goal.

“It took a little patience but in the process of providing free screening and at least showing parents how to navigate their way to the front of a line as more of an information service we were able to build the data models to support a development of a diagnostic device actually a couple of years sooner than we originally thought we would. So it ultimately paid off for us.”

Cognoa has raised around $11.6M in investor funding to date, according to CrunchBase, from the Chinese private investment group Morningside. Vaughan tells TechCrunch it’ll likely be looking to raise another round by the end of this year.

It has also conducted multiple studies over the last 2.5 years across the US, including blinded control trials and side-by-side comparisons of its different versions — working with children’s hospitals and secondary care centers. It now bills its technology as a “pediatric behavioral health diagnostics and digital therapeutics platform”.

The initial machine learning model, which was targeted at screening for autism, was based on the work of Stanford pediatrics and psychiatry professor Dennis Wall. The model itself was built by combining and structuring existing datasets of behavioral observations on about 10,000 children.

Though, as noted above, Cognoa has continued to refine its autism model with structured contributions from parents participating in the pilot and inputting data via its app. (Aka: If an AI service is free, you’re the training data.)

  1. Mockup - Assessment

  2. Mockup- Activity Details

  3. Cognoa app

  4. Mockup- Assessment Results

“In our last study we were able to come through with a sensitivity of greater than 90 per cent,” Vaughan tells TechCrunch. “In our first algorithm… targeting autism, we would find it over 90 per cent of the time — and when we said it was autism it was correct well over 80 per cent of the time.

“What we see when we look in the data, and that we’re quite interested by, is when we say it’s autism or it looks like autism and it wasn’t… we were able to show [the FDA] that they were often very similarly related conditions.”

Vaughan says a lot of the team’s early work focused on figuring out how to create a product that enables non-healthcare professionals (i.e. parents) to capture robust data in a reproducible way. “One of the… questions that came up quite early, even from early potential investors and clinicians, was can you actually get parents to give you the information on which you could base a clinical diagnostic decision? Can you get them to do this reproducibly without a clinician being in a room?… So we certainly had to address that.

I remember sitting down with one venture capitalist who looked at me and said, you know what — you’re never going to find 5,000 parents that are going to do this.

“I remember sitting down with one venture capitalist who looked at me and said, you know what — you’re never going to find 5,000 parents that are going to do this. And that are going to be able to do this reproducibly,” he continues. “Within a couple of years we were up over a quarter of a million parents that had actually done it — and we learned a lot about how to reproducibly collect information on which you can build a clinical diagnosis but collecting it outside of the clinical setting. Parents providing us information in their living room in the evening. So that was certainly one major step for us. And in doing that we showed that the unmet need was much, much bigger than we originally had estimated.”

As well as aiming to support earlier diagnosis than parents might be able to get if they had to wait for specialist appointments for their child to be monitored in person, Cognoa’s platform provides guidance on actions (it calls them “activities”) parents can take themselves to help manage their children’s condition. Which in turn provides more opportunities for response data to be fed back so its models can keep learning and refining recommendations.

While the first focus is autism, with the aim of trying to shrink intervention times to improve long term outcomes for children — given what Vaughan describes as a “well-documented” link between earlier intervention and better autism outcomes — the intent is to address other behavioral conditions too, in time, such as ADHD.

“For us we see this — even the autism clearance that we’re looking forward to in the future — that’s just a step down the path of being able to be the platform that can diagnose an entire spectrum of these developmental conditions,” he says.

Interestingly, Vaughan concedes that the learning element of AI-based technologies can cause unintended problems in healthcare service provision, saying some clinicians it talked to early on raised concerns that by widening access to autism screening the startup risked making an existing diagnosis bottleneck worse by increasing demand for specialist services without there being a parallel increase in resource to avoid creating even more of a backlog.

Which is exactly the kind of serious, knock-on consequence that’s possible when unproven ‘disruptive’ technologies change existing dynamics and bring new pressures to bear on a critical and sensitive industry like healthcare. It also seems especially true of AI technologies which need to be fed with lots of data before they can learn to become really useful.

So how to conduct responsible training of machine learning models presents something of an existential challenge for AI and healthcare startup initiatives — and one which has already opened up operational pitfalls for some very well resourced tech giants.

“Back in 2014 and 2015 we were really starting down the path of let’s just prove that we can triage these kids and find them earlier. And a lot of people embraced that, but there was certainly some that were pretty thoughtful who said if you guys find the kids earlier and the problem in the system is that kids that are identified and referred to specialists for appointments are currently waiting between one and three years to get a diagnosis, aren’t you just going to be making the problem worse?” he says.

“So then we had to sit down and say listen, step one is being able to show that we can just screen these kids. But longer term we think we can really aid in getting a faster diagnosis. But we were very careful to not say, publicly, that we thought that we could diagnose these kids because we thought it would just be too controversial. And the idea of using an AI-based platform, the idea of collecting information primarily from the parent, from the caregiver and from the child, that was pretty controversial.”

Another change that’s being driven by AI-based software targeting the healthcare industry is to regulatory regimes — with regulators like the FDA needing to come up with new systems and processes for assessing and managing software designed to get better over time.

“The FDA is struggling with how to regulate AI-based software because the idea of the FDA is they look at a version of a product and that product once cleared by the FDA does not change — and the idea of AI and machine learning, which is what our product is based on, is that it’s learning and it gets better,” says Vaughan, talking about its discussions with the regulator. “And so understanding with the FDA how we were going to control and document that learning — those were some of the discussions where we walked in with ideas but not very clear understanding what the outcome would be.”

While he believes the FDA will likely take a case-by-case approach to the challenge of regulating AI platforms, he suggests companies will probably have to operate using a versioning system — whereby they restrict ongoing learning to the research lab, releasing a next version of a model into the wild only once the step change in their model has also gained regulatory approval.

“It’s the algorithm part of the device that [the FDA] feel the strongest about in terms of how they regulate it,” he says. “And keep in mind this is evolving, and their thinking might also evolve on this, but for us they look at the algorithm part and we can certainly, in our software, lock down a current version of the algorithm. And we can allow that to not change in the production version of the product — and at the same time we can have a research arm that’s continuing to evolve. And you could start to think about versioning coming out in the future.”

“So I think it’ll be a little bit more of a stair-step approach,” he adds. “With periodic reviews by the FDA. And I think that they’re in parallel trying to think of a way to streamline that approach going forward because of the flexibility that these products have. So I think it’ll be a little bit of a hybrid between continuous machine learning which seems quite difficult and the old style, which was quite waterfall.”

Featured Image: Cognoa

Avro aims to deliver drugs to children and the elderly through skin patches


Avro, a life sciences startup in Y Combinator’s current batch, is banking on a method to deliver medications to populations unable to swallow or chew — It will administer them through the skin.

Starting with allergy medications, the startup is developing skin patches that release drugs commonly used in seasonal allergies for children. The patches act much like nicotine patches, which deliver nicotine to those trying to quit smoking, but can deliver a variety of drugs, such as allergy medications, through our body’s largest organ.

Co-founder Shakir Lakhani, who suffers from both seasonal and food allergies himself, tells TechCrunch he wanted to start with allergy relief not just for personal reasons but because its an area where children are resistant to taking the medications already out there.

“There’s medications that say they taste like banana but don’t really taste like banana, ” he said. “The drugs are also relatively safe so it’s something we can work with without being too worried.”

It’s also not the first company to offer transdermal drug delivery. Miami-based pharmaceutical company ProSolus creates skin patches for delivery of various generic drugs and over-the-counter products, Fremont outfit Zosano makes specialty transdermal patches for delivery of migraine-reducing medicines and Viaskin is making a patch for kids with peanut allergies — though clinical trials have so far not gone well.

These companies could easily start offering the same types of patches as Avro, should the startup prove the market need.

One advantage is that the drugs Avro wants to use in its patches are already out there — it wouldn’t need to prove they work. They just need to prove the delivery method is safe and effective.

But the startup has a long way to go before getting these patches into the hands of consumers. First, Avro will need FDA approval to sell in the United States and Canada, where the company plans to market its patches. To get there, it will need to conduct some human clinical trials, which Lakhani says he’s in the process of looking into at the moment but believes he’ll be able to get up and running in Q3 of this year.

Lakhani also mentioned his skin patches could eventually offer other types of medicines.

“We’re looking at things like people suffering from near degenerative diseases and more intense diseases that inhibit your ability to swallow like Multiple Sclerosis,” he told TechCrunch. “I think those will be really interesting avenues for us to go down in the future and we’re just starting to initiate conversations with other companies who might be good partners for us.”

RightEye’s portable eye-tracking test catches concussions and reading problems in five minutes

They say the eyes are the windows to the soul, but physiologically speaking, they’re really windows to the brain.

RightEye looks through that window to detect common but often subtle vision issues resulting from concussions and other brain troubles. Its quick, portable eye-tracking station can tell in minutes whether you should see a doctor — or look into becoming a pro ball player.

It turns out there’s quite a lot you can tell from how someone’s eyes move. We may not notice it ourselves, but we all vary in how and how well we execute a number of basic tasks, from flicking our eyes back and forth to smoothly tracking a moving target. For instance, your eyes may over-correct, fail to line up correctly, or track up or down when moving along a straight line.

For healthy individuals, these variations fall within a safe range, just part of the ordinary differences between bodies. But certain patterns well outside the baseline can be strong indicators of things like concussions and eye muscle problems — and even Parkinson’s and Autism-spectrum conditions.

RightEye tracks these movements with a custom device that looks a bit like an all-in-one desktop; it uses a Tobii eye-tracking module built into a single-purpose computer loaded with a library of simple tests. A basic EyeQ (as they call it) test takes five minutes or so, with more specialized tests adding only a few more, and results are available immediately.

To give you an idea: one test in game form has you defending a space station, destroying incoming ships by looking at them. But certain colored ships you must not destroy — meaning you have to detect them in your peripheral vision and avoid looking at them. In another test, you flick your eyes rapidly between two targets appearing on opposite sides of the screen, demonstrating accuracy and functioning saccades (micro-corrections made by your eye muscles).

Each eye is tracked independently, and their performance as a matched pair is evaluated instantly. An easy-to-understand results sheet shows their actual movements and how (if at all) they deviate from the baseline.

It’s compact and can run on battery for some 8 hours, making it ideal for deployment outside hospitals or the like: anywhere from school nurse’s office to the sidelines of an NFL game, even in the home.

[embedded content]

I tested the device out myself at CES (my vision is just OK, but I want a rematch), and later chatted with Barbara Barclay, RIghtEye’s President. The two most exciting applications of the technology, as judged by her enthusiasm anyway, are in identifying vision-related cognitive problems in kids and in creating a sort of eye fitness test for sporting persons.

Say a child is having trouble learning to read, or perhaps can’t pay attention in class. The immediate thought these days is frequently ADD. But it’s more than a little possible that it’s a vision problem. A subtle difference in how the eyes track, perhaps one going off the horizontal when tracking a line of text, could easily make reading on the page or blackboard frustrating or even painful. What 3rd-grade kid would keep at it?

A reading-focused test tracks how the eyes move along a line of text.

This isn’t some groundbreaking new idea — but reliably and objectively evaluating individual eye movements was only something you could do if you went to see a specialist, perhaps after other explanations for a behavior didn’t pan out. RightEye’s test practically runs itself and can detect or eliminate the possibility of vision problems in minutes. Honestly, I think a kid might even find it fun.

Barclay has personal experience with this, her own daughter having had health issues that only after multiple false starts were found to have their root in a relatively simple vision problem the system indicated.

In 2016, RightEye acquired the rights to a pair of tests based on research linking eye movement patterns to Parkinson’s and Huntington’s diseases, as well as Autism spectrum conditions. It’s not a magic bullet, but again, the quick and easy nature of the tests make them ideal for routine screening.

The Autism spectrum test is for children aged only 1 to 3, and watches eye movements between images of people and images of geometric shapes. Lingering on the shapes more than the people, it turns out, is a good indicator that at the very least the kid should receive further testing.

The Parkinson’s and Huntington’s tests watch for the more well understood patterns that accompany the motor degeneration found in those afflictions. They can be administered to people of any age and have (using earlier eye-tracking setups) contributed to many identifications of the diseases.

On a very different, but perhaps more immediately remunerative, note, Barclay told me that the test also works as a way to find outliers on the other end: people with what amounts to super-vision.

It’s entirely possible that someone could take the test and their results will show that they have faster, more accurate saccades, quicker target acquisition, and better continuous object tracking than the baseline. That’s a heck of an asset to have if you’re batting, fielding, goalkeeping, playing tight end — pretty much anything, really.

Examples of a healthy eye movement report (left) and concussed one (right).

It’s also a heck of an asset to have if you’re a scout or coach. If Lopez is catching great on the left side of the field but not the right, you can look into the possibility that he’s having trouble tracking the ball when looking over his left shoulder, his eyes all the way to the right.

Not only that, but you can test for effects of concussions or other traumas right there on the field if they’re having trouble. Given how widespread such injuries are and the immense danger of repeated concussions, testing early and often could literally save lives.

Right now, Barclay told me, 7 MLB teams are using RightEye tech for player assessments. As for the medical side of things, she said the company currently has 200 clients. The new hardware should help boost that number.

Perhaps more importantly, it has the backing (and therefore clout) of VSP, the country’s largest vision insurance company. That’s both a tremendous vote of confidence and a major in — nothing gets people using a system faster than knowing it’s covered by their existing insurance.

Swiss pharma company Roche is buying Flatiron Health for $1.9 billion


Roche, the global pharmaceutical company from Switzerland today announced it will scoop up Flatiron Health, a startup analyzing real-time oncology data to help cancer patients and doctors in a $1.9 billion deal.

Flatiron has also confirmed the deal in a tweet.

Two years ago, Roche led a $175 million deal in the startup at a $1.2 billion valuation. At the time of the deal, Roche agreed to buy several of Flatiron’s subscription-based software products, positioning the company for an eventual initial public offering.

Flatiron CEO and co-founder Nat Turner said back then he planned to IPO in “two to three years,” according to the New York Times. The plan was to raise yet another round of funding before doing so. However, it seems Roche has other plans.

“This is an important step in our personalised healthcare strategy for Roche, as we believe that regulatory-grade real-world evidence is a key ingredient to accelerate the development of, and access to, new cancer treatments,” Roche CEO Daniel O’Day said in a press release regarding the acquisition today. “As a leading technology company in oncology, Flatiron Health is best positioned to provide the technology and data analytics infrastructure needed not only for Roche, but for oncology research and development efforts across the entire industry.”

O’Day also mentioned the need to preserve Flatiron’s autonomy as a subsidiary. Turner also mentioned to CNBC that all employees, including the founding team, would stay on with the company.

Founding members Turner and Zach Weinberg both hailed from Google before pitching in to healthcare. In total, the two were able to raise over $313 million since Flatiron’s launch in 2012.

The $1.9 billion deal, which will be on a fully diluted basis and subject to certain conditions, is expected to close in the first half of this year, according to Flatiron.

Featured Image: Getty Images

Swiss pharma company Roche is buying Flatiron Health for $1.9 billion


Roche, the global pharmaceutical company from Switzerland today announced it will scoop up Flatiron Health, a startup analyzing real-time oncology data to help cancer patients and doctors in a $1.9 billion deal.

Flatiron has also confirmed the deal in a tweet.

Two years ago, Roche led a $175 million deal in the startup at a $1.2 billion valuation. At the time of the deal, Roche agreed to buy several of Flatiron’s subscription-based software products, positioning the company for an eventual initial public offering.

Flatiron CEO and co-founder Nat Turner said back then he planned to IPO in “two to three years,” according to the New York Times. The plan was to raise yet another round of funding before doing so. However, it seems Roche has other plans.

“This is an important step in our personalised healthcare strategy for Roche, as we believe that regulatory-grade real-world evidence is a key ingredient to accelerate the development of, and access to, new cancer treatments,” Roche CEO Daniel O’Day said in a press release regarding the acquisition today. “As a leading technology company in oncology, Flatiron Health is best positioned to provide the technology and data analytics infrastructure needed not only for Roche, but for oncology research and development efforts across the entire industry.”

O’Day also mentioned the need to preserve Flatiron’s autonomy as a subsidiary. Turner also mentioned to CNBC that all employees, including the founding team, would stay on with the company.

Founding members Turner and Zach Weinberg both hailed from Google before pitching in to healthcare. In total, the two were able to raise over $313 million since Flatiron’s launch in 2012.

The $1.9 billion deal, which will be on a fully diluted basis and subject to certain conditions, is expected to close in the first half of this year, according to Flatiron.

Featured Image: Getty Images

With $250 million, Peter Diamandis’ new startup is all about taking stem cells from placentas


Stem cells derived from a human placenta hold the key to unlocking a myriad of potentials in regenerative medicine and are the focus of X-Prize and Singularity University founder Peter Diamandis’ new endeavor.

Called Celularity, the startup is a spinout from Celgene, a global biopharmaceutical company creating gene therapies. Diamandis teamed up with Dr. Robert Hariri, the founder of Celgene, to create Celularity in the hopes of using stem cells found in the human placenta to quickly regenerate tissue and organs needed to treat cancer and other diseases. The idea is these types of cells can do a better job of helping us live longer, fuller, healther lives in the future.

It’s a wild proposal and, seemingly, the stuff of science fiction often tossed around in certain Silicon Valley circles — create a startup focused on a medical breakthrough to make us live forever — or at least much, much longer than we currently do. But stem cell technology has been around for some time.

Lab worker with human stem cells.

For decades stem cells have posed an ethical quandary as they’ve largely been harvested through discarded embryos. However, in just the last few years, science has discovered adult stem cells can come from a number of sources throughout the human body — including a woman’s placenta, shortly after giving birth.

Placental stem cells are even more important as they can be taken from any placenta and injected into any human without the risk of the body rejecting them, according to the company. And, because they are so abundant, treatments are potentially more affordable and can begin immediately.

But this is not the first time Diamandis has dipped his toe in longevity research. He cofounded Human Longevity Inc. in 2014 to focus on extending the human lifespan. Celularity extends his interests in this endeavor.

So far, the startup has conducted several clinical trials and treated “hundreds” of patients, Hariri tells TechCrunch. The next step is to try and gain FDA approval to roll these treatments out on a mass scale.

That approval may be just around the corner — possibly in the next 12-24 months, according to Hariri. That’s because “cellular medicine is intrinsically safe,” he says, adding the potential could have a “huge impact” on U.S. medicine.

So just how are Diamandis and Hariri obtaining these human placentas? Donations. Though some couples choose to keep (and later eat) their placental afterbirth, approximately four million human placentas are disposed of per year in U.S. hospitals.

While this strange, yet magical temporary organ is from a human mother, some states consider it to be a biohazard after birth and discard it as waste. Couples, by law, therefore cannot take and sell the tissue to Celularity or some other outfit hoping to use it up. But they can donate it.

That’s where Celularity comes in. The startup procures the placental tissue from hospitals willing to hand over would-be waste in the name of science.

Right now the field is pretty wide for Celularity, too. The company is only competing with a handful of others in the same space like Israeli biotech firm Pluristem. Though it seems to be the only player in the U.S. at the moment.

Add to that a whopping $250 million in new funding from several prominent investors and celebrities to help it grow, including well-known life coach Tony Robbins, John Sculley of Apple and Pepsi-Cola, former GV partner and founder Bill Maris, who now runs a new biotech funding venture Section 32, and Andrew Von Eschenbach, the former commissioner of the U.S. Food and Drug Administration.

Naturally, Celgene led the latest round, with cash infusions from United Therapeutics Corporation, Sorrento Therapeutics and Human Longevity, Inc. to boot.

One other big heaping help to the startup — Celularity owns and operates LifeBank USA, the world’s only repository of donated placenta cells and biomaterials. So, presumably, should any other companies want to do something similar in the U.S., they’d need to go through a subsidiary of Celularity first.

Celularity is sure to stay a large player in this field as it continues. It owns the whole chain from procurement to deploying treatments. It also holds 1,800 patents on the procedures.

“Our ultimate mission is to make 100 years old, the new 60, and to provide people with maximal aesthetic, mobility, and cognition as they age,” Diamandis said. “The 20 years of science, research, and intellectual property pioneered by my visionary partner Dr. Bob Hariri, is the cornerstone for the coming longevity revolution.”

Featured Image: BSIP / Contributor/Getty Images

Fitbit buys Twine Health in bid to become a more serious health care tool


Fitbit’s been on a bit of a acquisition spree over the last couple of years, as the company’s looked to grow its business inside the stagnating wearables category. This morning, the hardware maker announced plans to pick up Twine Health, a HIPAA-compliant, cloud-based health management platform.

The company hasn’t disclosed specific numbers for the acquisition, but expects the deal to close at some point in Q1 of this year. The integration plan seems pretty straight forward, leveraging Twine’s service with Fitbit’s large customer base, in an attempt to offer up more complex and useful health care data collected by its line of wearables. Specifically, the dataset will focus on hypertension and diabetes.

“When combined with our decade-plus of experience empowering millions of consumers to take control of their health and wellness,” Fitbit CEO James Park said in a press release issued this morning, “we believe we can help build stronger connections between people and their care teams by removing some of the most difficult barriers to behavior change. Together, we can help healthcare providers better support patients beyond the walls of the clinical environment, which can lead to better health outcomes and ultimately, lower medical costs.” 

Fitbit and Apple have been particularly aggressive in a push to get their wearables taken more seriously as potential health care tools. In recent years, fitness trackers and smartwatches have become among the most widespread personal monitoring devices around.

While the products don’t rate as sophisticated medical equipment, they do contain enough data tracking to provide potentially useful insight into a large cross section of the population. With this acquisition, Fitbit is no doubt hoping that its line of devices will become more indispensable for users with chronic health conditions that require daily tracking.

No specifics on what will happen to Twine’s Cambridge, Massachusetts-based staff, but the company is  set to roll into Fitbit’s Health Solutions wing, with co-founder and CEO John Moore serving as the company’s Medical Director.